548 CELL MECHANICS 



theses connecting observed structures of the chromosomes with 

 the occurrence of crossing-over between them, have supported 

 their views with hypotheses explaining what forces determined 

 the crossing-over. In some cases this mechanical proposal merely 

 added an ornament that was lacking in Janssen's original teleo- 

 logical construction. In others it was made the sole foundation of 

 the theory. In all it depended on the assumption that the coiling 

 of the chromosomes round one another, at the various stages at 

 which it was thought that crossing-over might occur, determined 

 its occurrence. 



Now, however, we are in a different position. On the one hand 

 we know the exact time and place of crossing-over. And on the 

 other hand the forces acting on the chromosomes at different 

 times can be distinguished and related according to their origins 

 as well as according to their effects. The paired chromosomes are 

 relationally coiled^ and crossing-over between two of their four 

 chromatids at the moment of their origin removes this coiling and 

 determines a chiasma (D., 1935 c and d). 



Crossing-over consists in a sequence of events, presumably a 

 rapid sequence, at the end of the pachytene stage of meiosis. Two 

 chromatids of partner chromosomes break, and they break at 

 exactly opposite points. The breakage of one must therefore 

 determine that of the other and precede it in time. The four ends 

 must then move in a way that allows them to come together 

 in a new combination. They may then rejoin to form two new 

 chromatids. This rejoining presents no new problem, for as we 

 have seen, ends of different chromosomes, free and unpaired, are 

 capable of rejoining after X-ray breakage. The problem consists 

 in the original breakages and the movement. 



The position of the paired chromosomes at the end of pachytene 

 can be exactly defined. Each particle is associated with a corres- 

 ponding particle of the partner ; it is undergoing division ; the two 

 threads are twisted round one another by relational coiling. The 

 forces at work can be similarly defined. The lateral association 

 is by a specific attraction between successive particles. The 

 relational coiling must depend on the action of a second force 

 apart from the specific attraction, a longitudinal cohesion. These 



